Parental mentalization refers to a parents’ capacity and interest to consider the individual experience and mental state underlying the behaviors of the child. Higher mentalization is considered a key aspect for parental sensitivity in interaction, fostering child’s socioemotional and self-regulatory development. Yet, previous studies have not examined the dynamic pathways through which the maternal mentalization may develop, nor their effects on child development. Thus, in the current person-oriented study, first, we identify distinct profiles and longitudinal trajectories of maternal mentalization from pregnancy to child’s 2 years of age. Second, we test how the profiles and trajectories associate with children’s internalizing and externalizing problems, social–emotional competence and effortful control at the age of 2 years. Third, we examine how the profiles and trajectories associate with contextual demographic and child related. The substudy was part of the FinnBrain Birth Cohort and included families from general population (n = 2,687). Mothers reported their parental reflective functioning (PRF) at late pregnancy, 6 months and 2 years of child’s age. Both mothers (n = 1,437) and fathers (n = 715) reported the developmental child outcomes at the child’s age of 2 years. Latent Profile Analysis and Latent Transition Analysis were used to identify PRF profiles and trajectories. The results showed decreasing heterogeneity in PRF from pregnancy to child’s age of 6 months and 2 years (i.e., four, three and two latent classes, respectively). Most mothers progressed towards high PRF over time. Second, the profiles and trajectories depicting high PRF associated with child high social–emotional competence at the age of 2 years, yet no clear positive effects were found on child’s problems and effortful control. The group of mixed PRF trajectories showed strongest association with child’s internalizing and externalizing problems. Finally, there were theoretically meaningful associations between the PRF trajectories and both the contextual (e.g., parity) and child related (e.g., infant temperament) factors. This was the first study to explore the early unfolding of maternal mentalization. The results are discussed in relation with the potential mechanisms accounting for child development and with the nature and limitations of self-reported parental mentalization.